JP5186367B2 - Memory migration system and method - Google Patents

Description

  The present invention relates to the field of computer memory. More specifically, the present invention relates to automatically controlled memory migration.

  Many different types of computing systems are in widespread use throughout the world. These computing systems include personal computers, servers, mainframes, and various stand-alone and embedded computing devices. There are randomly-extending client-server systems where applications and information are distributed across multiple PC networks, mainframes, and minicomputers. In a distributed system connected by a network, a user can access a large number of application programs, databases, network systems, operating systems, and mainframe applications. Computers provide individuals and companies with a host of software applications including word processing, spreadsheets, accounting, email, voice over Internet protocol telecommunications, and facsimile.

  Users of digital processors, such as computers, continue to demand increasingly high performance from such systems to handle increasingly complex and difficult tasks. Furthermore, processing speed has improved much faster than main memory access speed. As a result, cache memory or cache is often used in many such systems to increase performance in a relatively cost effective manner. Many modern computers also support “multitasking” or “multithreading,” where multiple programs or thread of programs are executed alternately in the execution pipeline of a digital processor. A common architecture for high performance single chip microprocessors is the reduced instruction set computer (RISC) architecture, which is characterized by a small simplified set of frequently used instructions for high speed execution. Thus, in a RISC architecture, a single compound instruction contains a small set of simple instructions that are executed very quickly in multiple steps. These steps are performed in an execution unit adapted to execute certain simple instructions. In a superscalar architecture, these execution units typically include load / store units, integer arithmetic / logical units, floating point arithmetic / logical units, and graphical logical units that operate in parallel. In a processor architecture, the operating system controls the operation of the processor and the peripheral components of the processor. The executable application program is stored on the computer hard drive. The computer processor causes the application program to execute in response to user input.

  Memory storage is a critical component of almost every software application that exists today. As the storage industry has matured, the number of storage options has increased significantly and the performance of data entry and data retrieval has greatly improved. Storage options include magnetic tape, hard disk drive (HDD), compact disk (CD), digital video disk (DVD), and other media. Within the hard disk drive category, there are abundant storage options that range from simple and less reliable independent HDDs to high performance redundant disk arrays. This complicates the virtual storage environment. Key metrics for disk arrays include access speed and redundancy. Complex virtual storage environments include storage architectures that extend beyond a single computing environment. Such environments include a storage area network (SAN). Each storage mechanism has an associated cost and service level, both of which can vary greatly.

  In practice, many application programs experience peak periods of use, resulting in more stringent service level requirements. These programs experience a period of relatively low usage during the off-peak period, along with low service level requirements during the resulting off-peak period. The application program developer must determine which storage mechanism is required by the application program based on expected usage, cost, service level requirements, and other requirements. For example, a backup / recovery system might use a high service level storage subsystem to store frequently used or critical files whose loss can be catastrophic. May require. Often, the expected needs are different from the actual needs that arise when the program is installed. Furthermore, storage needs may change over time.

  Currently, when the anticipated need of an application program is different from the actual need, the application program administrator requests the storage administrator to change the storage service. Such changes are destructive and can be used to physically move files from one storage subsystem to another, reclaim unused storage, and provide additional storage. Requires downtime. The need to change may require changes to the storage layout. Layout changes usually involve temporary file migrations to make it easier for alternative storage layouts to meet changing demands.

  An object of the present invention is to provide a system, method, and program for performing automatic migration of data between a plurality of memory devices.

  According to a first aspect, a method for automatic migration of data between a plurality of memory storage devices, wherein one of the plurality of memory storage devices for one or more application programs is provided. Or at least one of monitoring access to a plurality of stored application program data and determining one or more service level specifications of the one or more application programs. And determining a change in memory allocation for one or more data of the application program from at least one of the monitored accesses and all determined service level specifications. Depending on changes in memory allocation, multiple And a to automatically transfer data between multiple memory storage devices to achieve an allocation of data in a plurality between of the over-di device, a method is provided.

  According to a second aspect, an automatic migration memory storage system comprising a plurality of memory storage units each associated with a predeterminable performance characteristic, comprising one or more application programs At least one of a data store that stores one or more service level specifications and a monitoring component that monitors access of application program data stored in at least one of the memory storage units An analysis component that further comprises one and determines reallocation of data among the plurality of memory storage units from at least one of the monitored accesses and all service level specifications; In response to a reallocation, a memory storage unit Automatic migration memory storage system further comprising a transfer control component to transfer data to another memory storage unit is provided by.

  According to a third aspect, the computer monitors access of application program data stored in one or more of the plurality of memory storage devices for one or more application programs; A machine-accessible medium comprising instructions for performing an automatic migration, a memory for performing an operation including determining at least one of one or more service level specifications of one or more application programs The operation determines a change in memory allocation for one or more data of the application program from at least one of the monitored accesses and all determined service level specifications. Change of the determined memory allocation A program is further provided that further performs automatic transfer of data between the plurality of memory storage devices to achieve allocation of data between the plurality of storage devices. The

  Preferably, a form of automatically migrating files from one memory subsystem to another with little or no disruption of application program usage is provided.

  Systems, methods, and media for automatic migration of data between multiple memory storage devices are provided. One embodiment includes monitoring access of application program data stored in one or more of the plurality of memory storage devices for one or more application programs. The process also determines service level specifications for one or more application programs. The method includes determining a change in memory allocation for one or more data of the application program from the monitored access and service level specifications. Data is automatically transferred between the plurality of memory storage devices to achieve allocation of data among the plurality of storage devices according to the determined change in memory allocation.

  Embodiments include an automatic migration memory storage system that includes a plurality of memory storage units each associated with a predeterminable performance characteristic. A data store stores service level specifications for one or more application programs. A monitoring component monitors access to application program data stored in at least one of the memory storage units. An analysis component determines reallocation of data among multiple memory storage units from monitored access and service level specifications. A transfer control component transfers data from one memory storage unit to another in response to the determined reallocation.

  Another embodiment of the present invention provides a machine-accessible medium that includes instructions effective when executed by a data processing system to cause the system to perform a series of operations that perform memory and automatic migration. To do. This operation includes monitoring access to application program data stored in one or more of the plurality of memory storage devices for one or more application programs. This operation also includes determining a service level specification for one or more application programs. This operation also includes determining memory allocation changes for one or more data of the application program from the monitored access and service level specifications. Data is automatically transferred between the plurality of memory storage devices to achieve allocation of data among the plurality of storage devices according to the determined change in memory allocation.

  Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings.

  Similar symbols may indicate similar elements.

  The following is a detailed description of example embodiments of the invention illustrated in the accompanying drawings. The example embodiments are so detailed as to clearly convey the invention. However, the amount of detail provided is not intended to limit the anticipated variations of the embodiments, but conversely, the intent is that of the present invention as defined by the appended claims. It includes all modifications, equivalents, and alternatives falling within the spirit and scope. The detailed description below is designed to make such embodiments obvious to a person of ordinary skill in the art.

  Disclosed are systems, methods, and media for performing automatic migration of data between multiple memory devices. In one embodiment, memory access for application program data is monitored for each of the one or more application programs. Data can be stored in one or more of a plurality of memory storage devices, each having its own performance characteristics. Monitored access is evaluated to determine the optimal distribution of application program data, usually stored in files, among multiple memory storage devices. This evaluation takes into account the service level specification of each application program. Periodically, data is automatically transferred from one memory storage device to another, determined between available memory storage devices consistent with service level specifications. Optimal allocation can be achieved.

  FIG. 1 illustrates a digital system 116, such as a computer or server, implemented in accordance with one embodiment of the present invention. Digital system 116 includes a processor 100 that can operate according to basic input / output system (BIOS) code 104 and operating system (OS) code 106. The BIOS code and the OS code are stored in the memory 108. The BIOS code is typically stored in read only memory (ROM), and the OS code is typically stored on the hard drive of the digital system 116. The memory 108 also stores other programs for execution by the processor 100 and stores data 109. The digital system 116 includes a level 2 (L2) cache 102 located physically close to the multithreaded processor 100.

  The processor 100 includes an on-chip level 1 (L1) cache 190, instruction buffer 130, control circuitry 160, and execution unit 150. The L1 cache 190 receives and stores an instruction close to execution. The instruction buffer 130 forms an instruction queue and allows control over the order of instructions issued to the execution unit. The execution unit 150 performs the operation requested by the instruction. Execution unit 150 may include a load / store unit, an integer arithmetic / logical unit, a floating point arithmetic / logical unit, and a graphical logical unit. Each execution unit includes a stage that performs steps in the execution of instructions received from the instruction buffer 130. Control circuitry 160 controls instruction buffer 130 and execution unit 150. The control circuitry 160 also receives information related to control decisions from the execution unit 150. For example, the control circuitry 160 is notified in the case of a data cache miss in the execution pipeline.

  The digital system 116 typically includes a Trusted Platform Module, a memory controller, a random access memory (RAM), peripheral drivers, a system monitor, a keyboard, a color video monitor, and one or more flexible diskette drives. Other configurations not shown, such as one or more removable non-volatile media drives such as fixed disk hard drives, CD drives, and DVD drives, pointing devices such as mice, and network interface adapters Also includes elements and subsystems. Digital system 116 may include a personal computer, workstation, server, mainframe computer, notebook or laptop computer, desktop computer, or the like. The processor 100 can also communicate with the server 112 via the input / output device 110. Server 112 connects digital system 116 to other computers and server 114. Thus, the digital system 116 can be in a network of computers such as the Internet and / or a local intranet. In addition, the server 112 can control access to another memory 118 including tape drive storage, hard disk array, RAM, ROM, and the like.

  In one mode of operation of digital system 116, the L2 cache receives data and instructions from memory 108 that are expected to be processed in the pipeline of processor 100. The L2 cache 102 is high speed memory that is physically located near the processor 100 to achieve higher speeds. The L2 cache receives instructions from a plurality of instruction threads from the memory 108. Such an instruction may include a branch instruction. The L1 cache 190 is located within the processor and preferably contains data and instructions received from the L2 cache 102. Ideally, when a program instruction is approached when it is executed, that instruction is first passed to the L2 cache along with the data, if any, and then is almost imminent. Sometimes passed to the L1 cache.

  The execution unit 150 executes the instruction received from the L1 cache 190. Execution unit 150 may include a load / store unit, an integer arithmetic / logical unit, a floating point arithmetic / logical unit, and a graphical logical unit. Each of the units can be adapted to execute a specific set of instructions. Instructions can be submitted in parallel to different execution units for execution. In one embodiment, two execution units are used simultaneously to execute certain instructions. Data processed by the execution unit 150 can be stored in and accessed from integer register files and floating point register files (not shown). The data stored in these register files can come from or be transferred to the onboard L1 cache 190 or external cache or external memory. The processor can load data from a memory, such as an L1 cache, into a processor register by executing a load instruction. The processor can store data from a register into memory by executing a store instruction.

  Accordingly, the system of FIG. 1 can include a plurality of computers having the processor and memory just described connected in a network serviced by a server. The server facilitates and coordinates communication between computers in the network. Each computer has its own memory for storing code and files and data for executing its operating system, BIOS, and application programs. Computer memory includes read only memory (ROM), cache memory implemented in DRAM and SRAM, hard disk drives, CD drives, and DVD drives. The server also has its own memory and can control access to other memories such as tape drives and hard disk arrays. Each computer can store its own application program. Some application programs, such as databases, can reside in the server. Thus, each computer can access the same database stored on that server. In addition, each computer can access other memory by the server. Thus, there is a memory gradation with faster and more expensive memory closer to the computer and server processors and slower and less expensive memory far from the processors.

  In the course of operation of the system of FIG. 1, data is moved from one memory location to another. For example, data may be moved from the processor's local cache to hard drive system memory. Data is typically organized into files on the hard drive. Files can contain text, objects, computer code, and so on. Thus, several data files are stored on the computer hard drive. Other files can be stored on a CD or DVD. A backup memory device can also be removably connected to the computer to back up data from the hard drive. In larger systems, some data files can be backed up to tape accessed via server 112. Other magnetic or optical memory can be used.

  In general, memory access speed increases as one approaches the processor. Thus, cache memory access is faster than fast disk access, fast disk access is faster than remote disk access, and remote disk access is faster than offline storage access. . Conversely, memory decreases in cost as memory access speed decreases. Thus, the cache memory is more expensive than a high speed disk or the like. Also, the memory size increases with decreasing memory speed. Thus, for example, a high speed disk is larger than a cache memory. Thus, in general, frequently used or changed data preferably resides in a cache or high speed disk. Data that is used or changed less frequently can reside in the remote disk array. Data that must be archived over a period of time can be stored on tape by a tape drive system.

  Memory usage is an important component in software development. Software developers write software in anticipation of available memory and expected usage. For example, a software developer may expect that only some of the total number of files on a program CD will be used for normal installation. Thus, the setup program may load only a portion of the program file from the program CD onto the hard drive. When a user selects a feature that is not loaded on the hard drive, the user is prompted to insert a program CD to retrieve a file that implements the selected feature. Obviously, software developers may misunderstand the actual use of memory during the operation of the software. Thus, embodiments can automatically adapt memory allocation according to actual use.

  Often, data files created and manipulated by an application program are stored on the hard drive of the computer. These files contain data and instructions that can be loaded into the computer's cache for processing. In the course of operation, certain files become dormant, i.e. they are not used for a relatively long time. In that case, these files can be stored in a more distant memory, such as a disk array, that is less costly and has more available space. Files that are dormant in the disk array can then be transferred to tape storage. Accordingly, embodiments include a system that monitors memory usage, compiles usage statistics, and transfers files from one memory device to another according to usage.

  FIG. 2 illustrates an embodiment of a system 200 that monitors memory usage and dynamically allocates memory. In general, memory can be described and implemented in layers. For example, the memory can be cache memory, high speed disk, remote disk, and offline storage. The cache memory is close to the computer processor and can usually be implemented with high speed random access memory (RAM). High speed disk storage such as hard drives are local to the processor. A remote disk array can also be used as the third layer of memory, possibly with multiple computers accessing the array. The fourth layer or offline storage includes magnetic tape. Offline storage serves as an archive of data that must be preserved but rarely used.

  Although three layers of memory 202, 204, and 206 are shown, different numbers of layers can be supported. For example, memory layer # 1 can include a high speed disk, memory layer # 2 can include a remote disk, and memory layer # 3 can include a tape drive. Other layers can include cache memory, backup media, and the like. Data can be transferred from one layer to another. In one embodiment, data can be transferred between layers 1 and 2, between 1 and 3, and between 2 and 3. The transfer of data from one layer to another is controlled by the memory transfer controller 214. Memory transfer controller 214 is responsive to control signals from analysis module 212 or user interface 216. In response to these control signals, the memory transfer controller 214 instructs transfer from one memory layer to another.

  The memory migration system 200 further includes a usage monitor 208 that monitors the usage of each layer of memory and accumulates usage statistics. Multiple application programs may be running simultaneously on the computer or network of computers. Each application attempts to make its own memory usage. Thus, instructions and data may be accessed from a hard drive in tier 1 during the execution of another program. Some applications may transfer data on a very frequent basis between the hard drive and the computer processor cache. Some applications may do this on an infrequent basis. The usage monitor 208 of the system 200 monitors file access rates and provides data to the analysis module 212. The usage monitor 208 also monitors and supplies data relating to access time, file size, and the like.

  Usage statistics can include how often a file is accessed or modified. By using usage statistics, embodiments can dynamically and intelligently allocate access to various memory implementations available in the system. These statistics can indicate that a file is being accessed from a layer of memory frequently enough to justify transferring the file to a higher layer of memory. Conversely, usage statistics can indicate that a file is being accessed infrequently enough to justify storing the file in a lower layer of memory. Thus, a file can be allocated to a fast local disk at some point and to a remote disk at another point. Similarly, files allocated to tape during one part of the day can be allocated to a remote disk during another part of the day. By examining historical and current trends in usage, embodiments can efficiently and automatically reallocate memory. Accordingly, embodiments monitor and maintain historical information regarding file usage. Historical data can form a baseline against which current usage trends are compared. Examples of metrics that show trends include last access time, time of change, file size, length of time to perform an operation (read, write, change, etc.), and frequency of access over time. Embodiments also monitor where files are located, special access requirements, performance requirements, etc. that arise from application functional requirements.

  The analysis module 212 analyzes the usage data determined by the usage monitor 208. Thus, for example, the analysis module 212 must transfer a monitored application file to a tier 2 remote disk because the file is infrequently used or because the file has been dormant for a significant amount of time. You can decide that you have to. Accordingly, the analysis module 212 can include usage threshold parameters for each layer of memory. Usage below a given threshold prompts the analysis module 212 to send a control signal to the memory transfer controller 214 to direct the transfer of files from one layer of memory to a lower layer of memory. it can. The analysis module 212 can obtain analysis parameters via a user interface 216, which can be used with a keyboard (R) -like menu to allow control of user input and system 200 functions. , Mouse, and video monitor.

  Memory usage by a computer or computer network can vary due to various factors, including time of day. For example, a database program accessed by a network of computers may be used more frequently during the working hours of the day, but after working hours, its use is less demanding. Become. In fact, some programs demand memory access heavily during one part of the day, while other programs demand memory access heavily during another part of the day. Thus, embodiments allow memory to be allocated according to time-demand.

Memory usage may also be determined according to a service level agreement (SLA) provided for memory access allocation according to service level specifications. For example, one user may have a SLA to provide service level that gives priority higher than the second user with a different SL A memory access. Embodiments allow memory access allocation according to such agreements as well as reallocation according to SLA changes. Service level data store 210 stores a user, a computer terminal or information regarding the service level provided to the application program. For example, a service level for a given application can be given high priority during the hours from 2pm to 4pm. These service level specifications are entered into the service level data store 210 by the user interface 216. Thus, the service level data store 210, occasionally, when a change of service level specification time occurs, or a new service level is implemented, may be updated.

  From the analysis module 212, the latest service level specification can affect whether and when to notify the memory transfer controller 214 to direct memory migration of files and / or blocks of data. It can be used. Thus, in one embodiment, the analysis module 212 must perform which memory migration if memory migration occurs from a given level of service provided to a particular application during a specified time. Can be determined. The analysis module 212 then informs the memory transfer controller 214 to indicate the prescribed memory migration. Accordingly, the migration of files from one memory layer to another memory layer is stored in the service level data store 210 and sent to the analysis module 212 when the peak usage period begins or ends. Can be instructed according to.

Whether data is stored anywhere in the plurality of memory layers is a function of the service class, the service class includes a different set of service level specifications. Thus, one system can implement various service classes. A service class service level specifications may be dependent on various factors, including the class of information to be business considerations and process. Business considerations include issues such as availability, recoverability, security, performance, and value. Information classes can include email, decision support, data warehousing, audio / visual, and so on. The data can be in block units, object units, or file units in multiple formats such as binary and compressed bitmaps, and can include information or computer code. Thus, various service level specifications can be taken into account when making memory access allocation decisions.

  As previously noted, the user interface 216 allows user input of service level specifications to the service level data store 210 and allows user input of threshold and control parameters to the analysis module 212. . User interface 216 also allows a user to intervene to direct memory transfer controller 214 to transfer files from one memory layer to another. Thus, the system administrator can therefore intervene to allocate memory for the purpose indicated by the user. These user-instructed changes can be subject to automatic migration as determined by the analysis module 212, or may override the migration decisions of the analysis module 212 in some embodiments or modes of operation. it can.

  In one embodiment, the memory transfer controller 214 includes circuitry that generates or triggers a memory read signal or memory write signal in response to a control signal from the analysis module 212 or the user interface 216. Memory transfer controller 214 can trigger or issue write signals to transfer data from higher memory layers to lower memory layers, and transfer data from lower layers to higher layers The reading signal can be triggered or issued to Thus, for example, the analysis module 212 obtains from the service level data store 210 a program file currently stored in a higher layer based on the usage data obtained from the usage monitor 208. It can be determined that it can be stored in a lower layer consistent with the service level specification made. The memory transfer controller 214 can then trigger or issue a write signal to copy the file from the higher layer memory to the lower layer memory.

  The memory transfer controller 214 also includes circuitry to allow data access regardless of where the data has been migrated. For example, a file stored on a local hard drive is determined to be used infrequently and the analysis module 212 transfers the file to a remote disk under the specifications of a given service level agreement. Assume that it is determined that it is possible. When this transfer is complete, the data is placed with different addresses in different memory devices. Application program processes attempting to access the migrated data must be able to logically find this data at the new address. This can be achieved in several ways. For example, in one embodiment, a directory is created that maintains a map between a file name known to the application and a location in the memory device where data in the file corresponding to that file name is stored. This directory is updated each time file data is migrated. Thus, the application program can ascertain the current address of the data in memory from this directory. Thus, in this non-virtual embodiment, the application is informed about the actual current address of the migrated data.

  As another example, in a virtual storage system, the central logical switch 218 of the memory transfer controller 214 can be implemented to automatically create a logical connection between the requesting application and the migrated data. Thus, in one embodiment, the central logical switch 218 of the memory transfer controller 214 receives a memory access request from a process that is executing an application program. This memory access request includes a virtual address. The central logical switch maintains an up-to-date map between the virtual address and the actual address of the application program data. Thus, the received virtual address is mapped to the actual current address of the migrated data. Thus, in such an environment, the application program process requesting data access does not need to know the actual address of the data.

  FIG. 3 shows a flowchart 300 of an embodiment of usage statistics and service level specification analysis to determine memory migration. At element 302, the system obtains block, file, and object level usage statistics. These statistics are examined (element 304). The system determines whether the statistics have changed over time (element 306). Otherwise, logic flows to element 310. If the statistics change over time, the system determines whether the service class has changed (element 308). Recall that each service class is associated with its own set of service level specifications. If the service class has not changed over time, logic flows to element 310. If the service class has changed, logic flows to element 322 to determine a new storage layer for the data.

At element 310, the system checks service requirements , including availability, recoverability, security, and performance requirements. Next, the system determines the service requirements whether the changes over time (element 312). If the service requirement has not changed over time , logic flows to element 316. If the service requirements have changed over time, the system determines whether the service class has changed (element 314). Otherwise, logic flows to element 316. If the service class has changed, logic flows to element 322 to determine the new storage layer.

In element 316, the system, route management data arrangement; time requirements of the data arrangement; and events, including a policy that includes a trigger and an action, examines the Information Life-cycle Management (ILM) requirements. Next, the system determines whether the ILM requirements have changed over time (element 318). Otherwise, logic flows to element 326. If the ILM requirements have changed over time, the system determines whether the service class has changed (element 320). Otherwise, logic flows to element 326. If the service class has changed over time, logic flows to element 322 to determine the new storage. Element 322 determines the new storage layer. The system then migrates the data to the new layer (element 324). The system then waits for the specified time interval and the process continues at element 302.

  Some embodiments of the present invention are implemented as a program product for use with a computer system, such as the digital system 116 shown in FIG. The program product can be used on other computer systems or processors. One or more programs of this program product define the functionality of the embodiments (including the methods described herein) and can be included in various signal bearing media. Exemplary signal bearing media include: (i) information permanently stored on a non-writable storage medium (eg, a read-only memory device in a computer, such as a CD-ROM disk readable by a CD-ROM drive); (Ii) modifiable information stored on a writable storage medium (eg, a floppy disk, hard disk drive, or writable CD disk in a diskette drive), and (iii) wireless communication Including, but not limited to, information communicated to the computer via a communication medium such as a computer network or telephone network. The latter embodiment specifically includes information downloaded from the Internet and other networks. Such signal bearing media represent an embodiment of the present invention when carrying computer readable instructions that direct the functionality of the present invention.

  In general, the routines executed to implement an embodiment of the invention may be part of an operating system or a particular application, component, program, module, object, or sequence of instructions. The computer program of the present invention typically consists of a plurality of instructions that are converted by a native computer into a machine accessible format and thus into executable instructions. A program also consists of variables and data structures that either reside locally in the program or are found in memory or storage devices. Further, the various programs described later in this specification can be identified based on the applications for which they are implemented in particular embodiments of the invention. However, any of the specific program terms that follow are used for convenience only, and thus, the invention is any specific application that is identified and / or implied by such terms. It should be understood that it should not be limited to just use.

  Thus, another embodiment of the present invention provides a series of operations for automatic migration of data between a plurality of memory storage devices when executed in a data processing system. A machine accessible medium is provided that includes instructions useful for execution. This series of operations generally includes monitoring access of application program data stored in one or more of the plurality of memory storage devices for one or more application programs. This operation also includes determining a service level specification for one or more application programs. This operation further includes determining a change in memory allocation for one or more data of the application program from the monitored access and service level specifications. This operation further automatically automates data between multiple memory storage devices to achieve allocation of data among multiple of the multiple storage devices according to the determined change in memory allocation. Transfer. This operation may further include comparing a baseline set of memory storage specifications to actual memory usage. This operation may also include determining a change in memory allocation resulting from a change in service level specified for the application program. This operation may further include determining the frequency of access of the application program's data or determining the duration of the access time.

  Although the invention and some of its advantages have been described in detail with respect to certain embodiments, various modifications, substitutions, and the like can be made without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that modifications can be made. While embodiments of the invention may achieve multiple objectives, not all embodiments within the scope of the appended claims will achieve all objectives. Furthermore, it is not intended that the scope of the application be limited to the specific embodiments of the processes, machines, articles of manufacture, material configurations, means, methods, and steps described herein. As one of ordinary skill in the art readily appreciates from the disclosure of the present invention, does it exist that perform substantially the same function or achieve substantially the same result as the corresponding embodiments described herein? Later developed processes, machines, products, material configurations, means, methods, or steps may be utilized in accordance with the present invention. Accordingly, the appended claims are intended to include within their scope such processes, machines, manufacture, compositions of materials, means, methods, or steps.

FIG. 2 illustrates an embodiment of a digital system in a network with a processor in the digital system. FIG. 1 illustrates an embodiment of a system for automatic migration of data between multiple memory storage devices. FIG. 5 illustrates a flow diagram of an embodiment that achieves automatic migration of data between multiple memory storage devices.

Explanation of symbols

100 processor 102 level 2 (L2) cache 104 basic input / output system (BIOS) code 106 operating system (OS) code 108 memory 109 data 110 input / output device 112 server 114 other computer and server 116 digital system 118 separate memory 130 instruction buffer 150 execution unit 160 control network 190 level 1 (L1) cache 200 system 202 layer 204 layer 206 layer 208 usage monitor 210 service level data store 212 analysis module 214 memory transfer controller 216 user interface 218 central Logic switch 300 Flow diagram 302 Step to obtain block, file, and object level statistics 304 Checking information classification 306 Determining whether statistics have changed over time 308 Determining whether service class has changed 310 Checking service requirements 312 Changing service requirements over time Step 314 determining whether the service class has changed 316 checking the ILM requirement 318 determining whether the ILM requirement has changed over time 320 whether the service class has changed Step 322 step 322 step to determine a new storage layer 324 step to migrate data to the new layer 326 wait for a specified interval

Claims (4)

A method for automatically migrating data between a plurality of memory storage devices included in a computer , wherein the plurality of memory storage devices are divided into three memory layers, and The memory layer is a high-speed disk storage, the second memory layer is a remote disk, the third memory layer is offline storage, and the migration includes the first memory layer and the second memory. Between the first memory layer and the third memory layer, or between the second memory layer and the third memory layer, the computer comprising :
Access frequency of data stored in one or more of the plurality of memory storage devices, access time of the data, size of the data, or usage of time required to access the data Obtaining statistics,
Determining whether the usage statistics with time to change,
In response to the usage statistics changing over time,
Determining whether a service class has changed, wherein the computer can implement various service classes, each service class having a service level specification for one or more application programs; Said determining step associated with different sets of:
Determining a new memory layer in response to a change in the service class and transferring the data to the determined memory layer;
Determining whether service requirements have changed over time in response to the usage statistics not changing over time, the service requirements comprising availability, recoverability, security, and performance; The determining step, which is a requirement ;
In response to the service requirements changing over time,
Determining whether a service class has changed, wherein the computer can implement various service classes, each service class having a service level specification for one or more application programs; Said determining step associated with different sets of:
Determining a new memory layer in response to a change in the service class and transferring the data to the determined memory layer;
Determining whether an information lifecycle management (ILM) has changed over time in response to the service requirements not changing over time, the ILM managing route of data placement Said determination step comprising a policy including events, triggers and actions;
In response to the ILM changing over time,
Determining whether a service class has changed, wherein the computer can implement various service classes, each service class having a service level specification for one or more application programs; Said determining step associated with different sets of :
Performing a step of determining a new memory layer and transferring the data to the determined memory layer in response to a change in the service class. The method of claim 1 , wherein obtaining the usage statistics comprises obtaining the usage statistics at each memory layer. A computer program for automatically migrating data between a plurality of memory storage devices provided in a computer, wherein the plurality of memory storage devices are divided into three memory layers, The first memory layer is a high-speed disk storage, the second memory layer is a remote disk, the third memory layer is offline storage, and the migration is performed between the first memory layer and the first memory layer. Between the second memory layer, the first memory layer and the third memory layer, or between the second memory layer and the third memory layer. Item 3. The computer program that causes each step of the method according to Item 1 or 2 to be executed. A computer for automatically migrating data between a plurality of memory storage devices,
A processor;
A memory connected to the processor;
A plurality of memory storage devices , divided into three memory layers, wherein the first memory layer is a high speed disk storage, the second memory layer is a remote disk, A plurality of memory storage devices, wherein the memory layer is offline storage, and the migration is between the first memory layer and the second memory layer, the first memory layer Between the second memory layer and the third memory layer, or between the second memory layer and the third memory layer,
The computer, wherein the processor reads the computer program according to claim 3 into the memory and executes the computer program .

Images